Computer technology has had a tremendous impact on medical imaging; without computers modern radiological modalities like CT and MRI could not even exist. The interpretation of medical images, however, is still almost exclusively the work of humans. In the next decades this is expected to change. Computers will be used more often for image interpretation. This research area is called Computer-Aided Diagnosis (CAD). All major companies in medical imaging have research labs devoted to CAD and there is a growing number of smaller companies active in the area.

Automatic segmentation and quantification are closely related to CAD. Segmentation is needed to determine which parts of the image the computer needs to analyze and in many cases the accurate delineation of objects already yields clinically valuable information. Quantification is often the ultimate goal of radiological examinations, once the diagnosis has been made the physician needs to determine the extent and progression of a disease.

In 2001, the Image Sciences Institute established a CAD research group devoted to these issues. The goal of this group is two-fold:

• to develop new methodological tools for CAD that can be applied to computerized detection of abnormalities, to automatic segmentation, to quantification of disease progression, and to differential diagnosis of lesions;
• to implement CAD systems for various applications, evaluate their performance in clinical practice and to promote their use in the clinic.

Methodological research currently focuses on segmentation and feature extraction using statistical models of shape and appearance and builds upon techniques from pattern recognition. The application areas are chest, both plain radiography and multi-slice CT data, brain perfusion with multi-slice CT, the detection of calcifications in arteries, the analysis of retinal images, and bone age assessment in hand radiographs.

In various projects, the CAD research group collaborates with industrial partners and other academic research groups. Among these are:

• 3mensio Medical Imaging
• Delft University of Technology, Pattern Recognition Group,
• Eyecheck BV
• University of Chicago, Kurt Rossmann Laboratories for Radiologic Image Research
• University of Iowa, College of Engineering Imaging Group
• University of Iowa, Department of Ophthalmology and Visual Sciences
• Philips Medical Systems
• R2 Technology

Below is a list of people in the CAD group is given, and an overview of current research projects.

Personnel

Projects

Publications

For a complete list of publications of the CAD research group, click here.

Former personnel

• Marco Loog obtained his PhD degree in 2004 on Supervised dimensionality reduction and contextual pattern recognition in medical image processing. He is now an assistant professor at the Information and Communication Theory Group at Delft University of Technology.
• Meindert Niemeijer obtained his PhD degree in 2006 on Automatic detection of diabetic retinopathy in digital fundus photographs. He is now with the University of Iowa, Department of Ophthalmology and Visual Sciences.
• Michiel Schaap obtained his Master degree in 2005 on noise reduction in medical imaging. He is now with the Biomedical Imaging Group Rotterdam.
• Ingrid Sluimer obtained her PhD degree in 2005 on Automated image analysis of the pathological lung in CT. She is now with the Free University Amsterdam.
• Joes Staal obtained his PhD degree in 2004 on Detection of elongated structures in medical images. He is now with Optasia Medical close to Manchester, UK.
• Corne Hoogendoorn obtained his Master degree in 2006 on Establishing and Evaluating Landmark Correspondence in 3D Statistical Shape Modeling. He is now with the Computational Imaging Lab of Pompeu Fabra University in Barcelona, Spain.